Geology Reference
In-Depth Information
from coseismic shearing of the root system or
trunk can take many decades.
Ring-width studies along the San Andreas
Fault provide clear evidence for an earthquake
in 1812-13 and a less well-expressed earthquake
in 1857 (Fig. 6.24). Coincidentally, both of these
years also coincide with drought years in the
master chronology. In the regional master chro-
nology, however, ring widths bounce back to
average within a few years. In contrast, all of the
trees damaged by earthquakes show a slow
recovery over several decades; some never
recover fully to the previous rates of growth.
Such studies are particularly useful because they
serve to clarify the paleoseismic record. In 1812,
California was sparsely populated, and, due to
seismic damage in some coastal communities, it
was thought that the earthquake occurred on
some fault near the coast. The tree-ring data
show that it occurred instead along the San
Andreas Fault, and, when combined with trench
data from distant sites, these data help define
the extent of ground rupture and permit an esti-
mate of the magnitude of these 19th-century
earthquakes ( Jacoby
et al.
, 1988).
Rather than relying on trees that were directly
disturbed by earthquakes, innovative tree-ring
studies along the Alpine Fault in New Zealand
have relied on a process-response rationale to
date prehistoric earthquakes (Wells and Goff,
2006). As witnessed with many recent earth-
quakes, strong seismic shaking can induce thou-
sands of landslides that lead to large increases
in downstream sediment fluxes (Dadson
et al.
,
2004; Harp and Jibson, 1996; Meunier
et al.
,
2008). Following each major earthquake on the
Alpine Fault, the enhanced sediment flux to
nearby river mouths has been reworked by
powerful Pacific storms into a new dune ridge
that accretes successively seaward. Once
stabilized out of the surf zone, the dune ridge
has soon been colonized by rainforest, such that
a cohort of similarly aged old trees is found on
each ridge (Wells and Goff, 2006). In general,
the oldest tree on each ridge post-dates by 20-50
years the earthquake that is interpreted to have
precipitated dune formation. Overall, the tree-
ring record extends back 500 years and records
four large earthquakes with a precision similar
to (or better than) that derived from individual
radiocarbon ages for each earthquake.
Rockfalls
Earthquakes near rugged mountains trigger rock-
falls. Clouds of dust due to tumbling rocks envel-
oped hillslopes in southern California even
during the larger aftershocks (typically
M
L
∼
4-5)
of the 1994 Northridge earthquake (Harp
and Jibson, 1996). The 1999
M
w
=
7.6 Chi-Chi
earthquake in Taiwan triggered over 20 000 land-
slides (Dadson
et al.
, 2004). The effects of seismic
shaking are not confined to areas directly adja-
cent to a rupture. John Muir reported rockfalls in
the Sierra Nevada that were associated with the
1872 Owens Valley earthquake, which occurred
over 100 km away. In New Zealand, rockfalls
have occurred greater than 300 km from the epi-
center of large earthquakes (Bull
et al.
, 1994).
Consider for a moment a talus cone or other
accumulation of fallen blocks. Over time, a cone
accretes blocks that have tumbled as a result of
avalanches, freeze-thaw cycles, or other biologi-
cal or meteorological events that loosen and
dislodge blocks. These additions to the talus are
considered to be randomly distributed with
respect to time. Against this background of inter-
mittent, randomly spaced additions of blocks,
rockfalls triggered by earthquakes will add a
pulsed signal in which many tumbling blocks are
added simultaneously to the talus deposits.
Presumably, if we could date the time of addition
of each block exposed on the surface of a talus
cone, we would expect to discern a “spiky”
record of seismically triggered rockfalls poking
through a background of steadier talus growth
unrelated to earthquakes (Fig. 6.25). This inter-
pretation, of course, assumes that earthquakes,
as opposed to very large storms, avalanches, or
spontaneous rockfalls, are responsible for epi-
sodic addition of large numbers of blocks.
Avoidance of rockfall deposits fed by avalanche
chutes and selection of sites situated beneath
short, steep slopes of weakly consolidated boul-
dery material, such as young glacial moraines,
can enhance the likelihood that earthquakes
will have caused most of the major additions of
new blocks (Bull and Brandon, 1998).
